The Zwift ramp test raises your target power by 20 W every minute until you can no longer turn the pedals. It then sets your FTP at 75% of your best one-minute power from the test. If your best minute is 300 W, your FTP is 225 W. A 350 W best minute gives 263 W, and 400 W gives 300 W. The test is short and hard to pace badly, but the fixed 75% factor suits some riders better than others. Riders with a big anaerobic capacity usually come out higher than they would on a 20-minute test.
How the Zwift ramp test works
Zwift’s own description is brief. After a warm-up, resistance rises 20 W each minute until you stop, and “your FTP is your best one minute power multiplied by 0.75.” There is also a Ramp Test Lite that starts lower and rises 10 W per minute instead of 20.
TrainerRoad’s ramp test uses the same rule. Its support article says it takes 75% of your highest one-minute power, and that “every second matters,” so partly finishing the last step still lifts the result. Its FAQ adds that if you ride above target in the final minute, the result is reduced slightly. TrainerRoad also notes that it may apply more complicated calculations depending on how closely you followed the test format, so its result is not always exactly 75% of your best minute.
Practical points that follow from the protocol:
- There’s no pacing decision. You hold the target until you can’t. That is the main advantage over a 20-minute test, which many riders start too hard.
- Most of the test is easy. The early steps are well below threshold for most riders. Only the last few minutes are really hard.
- Your best minute is usually your last full step. Because power only goes up, the best rolling 60 seconds normally ends at or just before the point where you failed.
Ramp test FTP calculation chart
Computed with the Endurance Toolbox FTP calculator (ramp protocol: best one-minute power × 0.75, rounded to the nearest watt). Zones use the same Coggan boundaries as the power zones calculator; the power zones chart explains what each zone is for.
| Best 1-min power | Estimated FTP | Zone 2 Endurance | Zone 4 Threshold | Zone 5 VO2 max |
|---|---|---|---|---|
| 250 W | 188 W | 104–141 W | 170–197 W | 198–226 W |
| 275 W | 206 W | 114–155 W | 186–216 W | 217–247 W |
| 300 W | 225 W | 125–169 W | 204–236 W | 237–270 W |
| 325 W | 244 W | 135–183 W | 221–256 W | 257–293 W |
| 350 W | 263 W | 146–197 W | 238–276 W | 277–316 W |
| 375 W | 281 W | 156–211 W | 254–295 W | 296–337 W |
| 400 W | 300 W | 166–225 W | 271–315 W | 316–360 W |
| 425 W | 319 W | 176–239 W | 288–335 W | 336–383 W |
| 450 W | 338 W | 187–254 W | 305–355 W | 356–406 W |
Every 20 W step you survive adds 15 W to your FTP. That’s why grinding out an extra 30 seconds at the end moves the result so much.
Ramp test vs 20-minute test vs 2×8: the same rider, three answers
Each protocol makes its own assumption about how much of your test power is “above threshold”:
- 20-minute test: average power × 0.95, the most common convention.
- Ramp: best minute × 0.75, per Zwift and TrainerRoad.
- CTS field test: the higher of two 8-minute efforts × 0.90, per CTS.
Shorter tests rely more on anaerobic capacity, the limited store of work you can do above your sustainable level. In the critical power model that store is called W′.
To show this, we modeled three riders with the same sustainable power (critical power, CP = 250 W) and different W′. We used the two-parameter critical power model (power = CP + W′ / time), the same model behind the critical power calculator. The 20-minute and 8-minute powers come straight from that equation. For the ramp, we simulated 1-minute steps starting at 100 W and rising 20 W, with failure when W′ runs out, then took the best rolling 60 seconds. Each test result was then converted with the site’s FTP calculator.
| Rider (CP 250 W) | W′ | Ramp best minute | Ramp FTP | 20-min avg | 20-min FTP | Best 8-min | 2×8 FTP |
|---|---|---|---|---|---|---|---|
| Diesel | 12 kJ | 329 W | 247 W | 260 W | 247 W | 275 W | 248 W |
| All-rounder | 18 kJ | 349 W | 262 W | 265 W | 252 W | 288 W | 259 W |
| Punchy | 26 kJ | 371 W | 278 W | 272 W | 258 W | 304 W | 274 W |
How to read it:
- For the diesel, all three protocols agree within 1 W. A small anaerobic store means each multiplier is close to removing the “above threshold” part correctly.
- For the punchy rider, the ramp says 278 W and the 20-minute test says 258 W. Both riders have the same sustainable power, yet the ramp gives the punchy rider 31 W more than the diesel. The 20-minute test gives only 11 W more. The shorter the test, the more W′ inflates the result.
- The 2×8 sits in between, as you would expect from its length.
The same simulation with 10 W steps, the step size of Ramp Test Lite, gives ramp FTPs of 231, 242 and 254 W for the three riders. A slower ramp spends more minutes above CP before failure, so W′ is spread thinner and the best minute ends up lower. The 75% factor then produces a lower number. Zwift pitches Lite at smaller or less experienced riders. Don’t compare results between the two ramp versions. That fits with lab work showing that protocol design changes the result: Michalik et al. (2019) found peak power in a continuous ramp was about 18 W higher than in a 3-minute step protocol for the same cyclists.
This is a model, not a measurement. It ignores pacing, motivation, cadence, and the fact that real fatigue in a ramp doesn’t work like a simple fuel tank. Treat the direction of the differences as the point, not the exact watts.
Is the ramp test accurate?
The short answer: it is a consistent way to estimate FTP, not a precise one, and its error depends on your physiology.
- We are not aware of a peer-reviewed study that validates the 75% ramp factor itself. The factor comes from the training platforms, not from a published lab comparison.
- The 20-minute test isn’t a gold standard either. Inglis et al. (2020) found power at maximal lactate steady state averaged only about 93% of the 95%-adjusted 20-minute FTP in well-trained riders, with large individual variability. The 20-minute value also failed to show a training-related improvement that the lab measure detected.
- FTP and critical power are close but not interchangeable. Karsten et al. (2021) found CP averaged 7 W higher than 20-minute FTP, with limits of agreement from −19 to +33 W.
Who the ramp tends to overestimate
Riders with strong sprint or punch: track riders, criterium specialists, cyclocross racers, and athletes from sports with a lot of short, explosive efforts. A large anaerobic capacity raises the best minute without raising sustainable power. The symptom is threshold intervals at ramp-derived Zone 4 that you can’t finish, even when you’re rested and fueled.
Who it may underestimate
Riders whose sustainable power is a high fraction of their maximum: long-course triathletes, time trialists, and very aerobically trained riders with little top-end. Their best minute is modest, but they can hold close to it for a long time. The symptom is “threshold” intervals that feel like tempo and a 1-hour effort above the ramp FTP.
Worked example 1: Converting a ramp result into zones
A 72 kg (159 lb) rider finishes a Zwift ramp test with a best minute of 312 W.
- FTP = 312 × 0.75 = 234 W, which is 3.25 W/kg.
- Zones from 234 W: Zone 2 130–176 W, Zone 3 177–211 W, Zone 4 212–246 W, Zone 5 247–281 W.
A sensible first week would include a long ride capped near 176 W and a 3×10-minute threshold session at 215–235 W. If those 10-minute blocks finish under control with steady breathing, the estimate is probably reasonable. If the third block falls apart, the ramp may have overestimated this rider.
Worked example 2: Ramp says 263 W, 20-minute test says 234 W
A punchy rider records a 350 W best minute on the ramp (FTP 263 W). Two weeks later they average 246 W for a well-paced 20-minute test (FTP 234 W). That’s a 29 W gap, and it moves the whole threshold zone:
| Source | FTP | Zone 4 Threshold | Zone 5 VO2 max |
|---|---|---|---|
| Ramp (350 W × 0.75) | 263 W | 238–276 W | 277–316 W |
| 20-min (246 W × 0.95) | 234 W | 212–246 W | 247–281 W |
The bottom of the ramp-based threshold zone (238 W) is close to this rider’s entire 20-minute average. Threshold work at ramp zones would mean riding near 20-minute race effort for every interval. That’s a typical overestimate from a high anaerobic capacity. The 20-minute number is the better training anchor for this rider. Their ramp tests are still useful for tracking change, as long as they compare ramp to ramp.
Limitations
- Compare like with like. Ramp, Ramp Lite, 20-minute and 2×8 results aren’t interchangeable. Track trends within one protocol.
- Indoor-specific. The ramp is ridden on a trainer. Smart trainer accuracy, calibration and indoor heat all affect the number. A trainer that reads high gives you inflated zones on that trainer and wrong ones on a power meter outdoors.
- ERG mode and cadence. TrainerRoad says ERG mode gives no advantage. In practice, though, how you fail (cadence collapse vs choosing to stop) changes how much of the final step you complete.
- Readiness. Fatigue, poor sleep or an underfueled day can lower any test, but a short maximal test is especially sensitive to motivation.
- Health. A ramp test is a maximal effort. If you have a heart condition or other medical concern, ask a clinician before doing one.
Run your numbers
- In the FTP calculator, choose Ramp Test and enter your best one-minute power from the test. Add body mass for W/kg. You’ll get the 0.75 estimate and a seven-zone table. To compare protocols as in example 2, switch to 20-Minute Test or 2×8 Minute Test, enter that result, and put your ramp FTP in Previous FTP to see the gap.
- Take the FTP you trust into the power zones calculator for a clean zone table to set up workouts.
- If you suspect a big anaerobic capacity is skewing your ramp, enter 3- and 12-minute best efforts into the critical power calculator. It estimates CP and W′ directly. A high W′ alongside a ramp FTP well above your CP is a sign that the 75% factor is flattering you.